Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2602.13967 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915799493509120 |
|---|---|
| author | Zhang, Ruicheng Li, Xinyi Xu, Tianyi Zhang, Shuhao Liao, Xiaofei Jin, Hai |
| author_facet | Zhang, Ruicheng Li, Xinyi Xu, Tianyi Zhang, Shuhao Liao, Xiaofei Jin, Hai |
| contents | Most evaluations of External Memory Module assume a static setting: memory is built offline and queried at a fixed state. In practice, memory is streaming: new facts arrive continuously, insertions interleave with retrievals, and the memory state evolves while the model is serving queries. In this regime, accuracy and cost are governed by the full memory lifecycle, which encompasses the ingestion, maintenance, retrieval, and integration of information into generation. We present Neuromem, a scalable testbed that benchmarks External Memory Modules under an interleaved insertion-and-retrieval protocol and decomposes its lifecycle into five dimensions including memory data structure, normalization strategy, consolidation policy, query formulation strategy, and context integration mechanism. Using three representative datasets LOCOMO, LONGMEMEVAL, and MEMORYAGENTBENCH, Neuromem evaluates interchangeable variants within a shared serving stack, reporting token-level F1 and insertion/retrieval latency. Overall, we observe that performance typically degrades as memory grows across rounds, and time-related queries remain the most challenging category. The memory data structure largely determines the attainable quality frontier, while aggressive compression and generative integration mechanisms mostly shift cost between insertion and retrieval with limited accuracy gain. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_13967 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Neuromem: A Granular Decomposition of the Streaming Lifecycle in External Memory for LLMs Zhang, Ruicheng Li, Xinyi Xu, Tianyi Zhang, Shuhao Liao, Xiaofei Jin, Hai Artificial Intelligence Computation and Language I.2.7; H.3.3 Most evaluations of External Memory Module assume a static setting: memory is built offline and queried at a fixed state. In practice, memory is streaming: new facts arrive continuously, insertions interleave with retrievals, and the memory state evolves while the model is serving queries. In this regime, accuracy and cost are governed by the full memory lifecycle, which encompasses the ingestion, maintenance, retrieval, and integration of information into generation. We present Neuromem, a scalable testbed that benchmarks External Memory Modules under an interleaved insertion-and-retrieval protocol and decomposes its lifecycle into five dimensions including memory data structure, normalization strategy, consolidation policy, query formulation strategy, and context integration mechanism. Using three representative datasets LOCOMO, LONGMEMEVAL, and MEMORYAGENTBENCH, Neuromem evaluates interchangeable variants within a shared serving stack, reporting token-level F1 and insertion/retrieval latency. Overall, we observe that performance typically degrades as memory grows across rounds, and time-related queries remain the most challenging category. The memory data structure largely determines the attainable quality frontier, while aggressive compression and generative integration mechanisms mostly shift cost between insertion and retrieval with limited accuracy gain. |
| title | Neuromem: A Granular Decomposition of the Streaming Lifecycle in External Memory for LLMs |
| topic | Artificial Intelligence Computation and Language I.2.7; H.3.3 |
| url | https://arxiv.org/abs/2602.13967 |